Group Adaptive scheduler

Example command:

python3 main.py run test/with_prior_preference_random/config.yaml test/with_prior_preference_random/jobs.xml test/with_prior_preference_random/experiment.xml -s GroupAdaptiveExtend -e GroupGradient -ep estimation_input -eo estimation_output -jtp 8 -rr False -wl 20 -en ex_name> log_name.log 2>&1 &

Available parameters:

• config.yaml, jobs.xml, experiment.xml point to the experiment configuration file, job configuration file and the experiment queue file, in that order

• -s : name of the scheduler algorithm [RoundRobin, Adaptive, GroupAdaptive, GroupAdaptiveExtend]

• -e : estimation algorithm [EpsilonGreedy, Gradient]

• -ep : input preference data folder

• -eo : output preference data folder

• -jtp : jobs_to_peek parameter - number of jobs to consider to schedule in each round

• -wl : waiting_limit parameter - for considering late job - only used with GroupAdaptiveExtend scheduler

• -rr : activate random arrival rate